Author: Jaime Allen, Juan Carlos Muñoz, Juan de Dios Ortúzar (2019)
Journal: Transportation Springer
Keywords: Transit satisfaction, Public transport, Objective attributes, Multiannual surveys, SEM-MIMIC, Heterogeneity
Public transport (PT) providers aim to ofer services that meet users’ satisfaction, and for this, they can control some operational service attributes such as frequency, speed, crowdedness and reliability. Understanding how these objective attributes afect user satisfaction is essential to improve it cost-efectively, but these associations have not been examined enough in the PT literature. This study aims to unveil how key transit operational variables actually experienced by users afect their satisfaction. We analysed data derived from a multiannual consumer satisfaction survey for the Santiago de Chile Metro system; between January 2013 and June 2016 (n=41,993), where approximately 1000 questionnaires were completed each month. We also gained access to a set of operational variables managed by Metro for the same period, including more than 1.4 million records. With this unique dataset, we frst developed a structural equation model (SEM) with users’ perceived attributes, fnding that safety, ease of boarding, response to critical incidents (CI), the number and type of CI endured, and information, were the variables that mostly afected satisfaction. We also examined heterogeneity in transit satisfaction with SEM-MIMIC models, by characterising the user population through their trip and socioeconomic characteristics, fnding a striking result: that as users age they are more satisfed with the system. Next, we assessed whether including operational service attributes, such as crowding levels, frequency, commercial speed and CI, added predictive power to the proposed model. We found that the number of CI, speed, frequency and crowdedness, plus their variability (measured through the coefcient of variation), afected transit satisfaction at signifcant levels. Including these objective service attributes provided more explanatory power to the SEM-MIMIC transit satisfaction models. Policy recommendations for improving satisfaction, derived from our results, are: to implement an automatic control system for the number of passengers on Metro platforms (as safety and ease of boarding are critical issues for passengers); and to deploy a comprehensive tactical plan to address CI: determine which happen more often, take actions to minimise them and provide better responsive actions.